Seeking balance: The role of Article 6(5) of Additional Protocol II in balancing justice and peace in transitions from armed conflict to peace

Abstract

This article explores the role of Article 6(5) of Additional Protocol II to the Geneva Conventions in balancing justice and peace during transitions from armed conflict to peace. It argues that the provision, which encourages the granting of broad amnesties at the cessation of hostilities, requires a re-evaluation of the obligation to investigate, prosecute and punish under international law. By analyzing the legal context and scope of Article 6(5), as well as its application in transitional justice models such as in Colombia, the article highlights how the principle of peace can be prioritized alongside justice without undermining victims’ rights. The discussion critically addresses maximalist interpretations of justice, presenting a nuanced approach that foregrounds restorative justice and the importance of reconciliation in post-conflict societies.

Developments in international humanitarian law (IHL) have traditionally focused on how hostilities are conducted (jus in bello). More recently, however, the role of IHL after hostilities (jus post bellum) and its impact on accountability, particularly in prosecuting crimes committed during hostilities, has received some attention.¹ An essential aspect of this discussion is the scope of Article 6 of Additional Protocol II to the Geneva Conventions (AP II),² which addresses “penal prosecutions”. In particular, Article 6(5) explicitly requires the authorities in power to “endeavour to grant the broadest possible amnesty to persons who have participated in the armed conflict, or those deprived of their liberty for reasons related to the armed conflict, whether they are interned or detained”. This makes it the only norm derived from an international treaty that explicitly addresses amnesties.

This article will argue that Article 6(5) of AP II significantly affects how the obligation to investigate, prosecute and punish is interpreted during transitions from armed conflict to peace. Because the provision prioritizes the principle of peace at the end of hostilities, it requires moving away from a maximalist interpretation that calls for the investigation, prosecution and punishment of all those responsible when massive human rights violations or grave breaches to IHL have occurred. This maximalist interpretation would mean equating justice, which is a principle (and so, like any principle, must be balanced with others), with a rule (which only allows for “all-or-nothing” interpretations).³ To support this argument, the article will (1) address the regulatory context, scope and role of Article 6(5) in balancing justice and peace; (2) provide a critical response to positions that advocate for a restricted application of Article 6(5); (3) discuss the ambiguities and maximalist positions in how the Inter-American Court of Human Rights (IACtHR) applies Article 6(5) that have an impact on interpretation trends in Latin America; and (4) examine the role of IHL and Article 6(5) in the framework of peace agreements, in light of the case of Colombia.

The role of Article 6(5) in balancing justice and peace: Regulatory context and scope during transitions from non-international armed conflicts to peace

Although the collective imagination may suggest otherwise, peace is a central purpose of IHL. Indeed, “by recognizing a minimum applicable standard, a minimum ethical rationality, international humanitarian law facilitates reciprocal recognition of the actors in conflict, and thus favours the search for peace and reconciliation in societies fractured by armed conflicts”.⁴ In this sense, the Geneva Conventions and their Additional Protocols were adopted to also protect the “potential victims” of armed conflict.⁵ Thus, IHL explicitly recognizes peaceful aspirations.⁶

Likewise, although one may think initially that IHL is not concerned with justice either, this is, in fact, also one of its purposes, as is clear from various conventional and customary rules. Thus, Article 3 common to the four Geneva Conventions (common Article 3) prohibits, in paragraph 1(d), “the passing of sentences and the carrying out of executions without previous judgment pronounced by a regularly constituted court, affording all the judicial guarantees which are recognized as indispensable by civilized peoples”.

In turn, Article 6 of AP II establishes the criteria for guaranteeing due process in penal prosecutions. For its part, Rule 139 of the International Committee of the Red Cross (ICRC) Customary Law Study contains a number of additional rules that develop the general principle that the parties to the conflict must “ensure respect” for IHL.⁷ Hence, from the interpretation of the obligation to ensure respect derived from common Article 1,⁸ as well as from Rule 151 onwards of the ICRC Customary Law Study,⁹ several rules pertain to individual criminal responsibility, command responsibility in investigations, and the commission and investigation of war crimes. This broader context, of course, affects the scope of Article 6(5) of AP II.

As this article will demonstrate, peace and justice are cross-cutting principles of IHL that directly affect the interpretation of the regulatory context and the scope and limits of Article 6(5) of AP II.

Regulatory context of Article 6(5) of AP II: Non-controversial issues

The regulatory context with respect to Article 6(5) of AP II features a number of uncontroversial issues. First, this regulation derives from the first (and so far the only) multilateral treaty devoted exclusively to regulating the application of IHL to non-international armed conflicts (NIACs), in developing common Article 3.

Second, Article 6(5) forms part of a broader provision (Article 6) which regulates, in general, the prosecution and punishment of criminal offences committed in connection with the armed conflict. Third, it remains in force as a customary rule: the ICRC itself and the specialized doctrine¹⁰ have affirmed that this regulation is accepted as international custom. In addition, this regulation is repeatedly cited by various international bodies,¹¹ courts¹² and national parliaments,¹³ the International Criminal Court (ICC),¹⁴ the European system¹⁵ and the Inter-American System of Human Rights.¹⁶

Fourth, Article 6(5) was included in the first draft that the ICRC sent to the conference leading to the adoption of AP II,¹⁷ and remained substantially unchanged (with broad consensus among States) in subsequent discussions – so much so that no State objected to including the amnesty concept as one conducive to the cessation of hostilities.¹⁸ In fact, the debates that arose focused on whether amnesty should be a mandatory or optional tool for States, and on who should grant it.¹⁹ This is particularly noteworthy given that the final text, while not proposing amnesty as mandatory, does use the expression “shall endeavour” (“s'efforceront”) to denote that the conference of States was inviting the parties to approve amnesties in these contexts, and not to discourage them.

Fifth, although Article 6(5) expressly refers to “amnesties”, this does not mean that it prohibits other analogous legal concepts that have more limited effects on investigation, prosecution and punishment, such as pardons, a decision not to pursue criminal prosecution, and suspended sentences.²⁰ In its commentary on Article 6(5), the ICRC established that amnesty is an “act by the legislative power which eliminates the consequences of certain punishable offences, stops prosecutions and quashes convictions”.²¹ It also established that it would not have been necessary to expressly include free pardon in the regulation, as all legislations provide for the possibility of free pardon, and that the regulation only prevents the execution of the penalty.²²

This regulatory context, as will be analyzed below, has specific legal effects on the scope of the obligation to investigate, prosecute and punish in contexts of transition from armed conflict to peace.

Impact of Article 6(5) on the obligation to investigate, prosecute and punish

As is made evident in the previous section, Article 6(5) has a role to play in prosecution and punishment as it creates at least one obligation on states to act (using the expression “shall endeavour”) and refers to granting the “broadest possible amnesty”. However, this obligation is at odds with developments in international human rights law (IHRL), and even with some emerging trends in international criminal law. As will be discussed, these include prohibitions and disincentives for States to grant amnesties because – under a prevailing vision of the principle of justice and the right of access to justice – the concept may violate victims’ rights. However, the IHL perspective analyzes amnesties as a concept with broader purposes, including other rights of the victims, especially related to preventing the conduct from recurring,²³ and the principle and value of peace.

Therefore, and given these special purposes of amnesty and similar concepts in IHL, this text argues that the victims’ right of access to justice, in the context of the end of hostilities, must be interpreted in light of IHL and particularly of the obligation arising from Article 6(5) of AP II. Specifically, this interpretation requires moving away from a vision that tends to equate the duty to investigate, prosecute and punish in contexts of transition from armed conflict to peace with an “all-or-nothing” rule. In this sense, a balance between the principles of justice and peace is required, giving special consideration to the weight that Article 6(5) places on peace, which should have undeniable effects on the balance to be achieved.

Some might think that this position implicitly undermines victims’ rights; however, the opposite is true. If conceived appropriately and with respect for the existing inviolable limits, this interpretation has the potential to achieve better protection of the human person. Indeed, as part of transitional justice models, amnesty and other similar concepts have the potential to contribute to access to justice from a broader perspective.²⁴ This is true for a number of reasons: first, because they have the capacity to adequately balance the paradigms of retributive justice and restorative justice.²⁵ The latter does not focus exclusively on the offender, but gives greater weight to victims and communities in balancing the equation.²⁶

Second, when accompanied by system investigation methodologies,²⁷ amnesties make it possible to uncover the complex criminal structures behind the crimes committed and reveal the patterns, complex forms of participation and liability of those most responsible.²⁸ As such, granting benefits and not focusing on the individual criminal liability of non-determinant participants incentivizes them to help clarify broader patterns of responsibility.²⁹ And third, these concepts could better guarantee the rights to truth and reparation, as when amnesties are conditional, the requirements for granting penal advantages include duties (ideally demanding) on beneficiaries to help guarantee these victims’ rights.³⁰ In this regard, McCold, Llewellyn and Van Ness note that alternative approaches to traditional justice have emerged in large part from the weaknesses and limitations of conventional trials and punishments, and that a deeper vision of peace involves several interconnected activities which include establishing the truth of what happened, punishing those most directly responsible for human suffering, and providing reparations to victims.³¹

The key, therefore, is in the conditions for accessing amnesties, and in how these conditions guarantee victims’ rights without sacrificing the principle and goal of peace. If the conditions for accessing amnesties are robust and demanding, this will also mitigate some of the risks highlighted by authors such as Mégret, who emphasizes that the promise of amnesties can create perverse incentives for parties to increase the level of violence in an armed conflict.³²

Finally, it is important to specify that the predominant role of IHL in the discussion does not mean that there is no dialogue between other international law regimes and IHL, especially in defining the limits and margins for designing transitional justice models. Indeed, there is a broad debate in the doctrine as to what these minimums, derived from international law, should be when designing and implementing justice models in political transitions.³³ However, no studies to date have attempted to identify and systematize these minimums with respect to transitions from armed conflict to peace,³⁴ the type of transition on which this article focuses.

In fact, in general, the literature and international and national judicial decisions have assumed that analysis of the limits does not differ according to the type of transition.³⁵ It is particularly striking how little importance the analysis has given to Article 6(5) of AP II as a tool that encourages the granting of amnesties in these contexts.³⁶ In this sense, while some might consider that the position proposed in this article goes against the tide, especially in relation to developments in IHRL and international criminal law, the fact is that there are still important debates, gaps and opportunities for development in the various international law regimes with regard to this important issue.

In this framework, and as argued in this article, a maximalist interpretation that makes the obligation to investigate, prosecute and punish an all-or-nothing rule (as has emerged with great vigour in the field of IHRL by demanding the investigation, prosecution and punishment of all the facts and all those responsible³⁷) is incompatible with the vision proposed in Article 6(5) of AP II. In light of this vision, legal tools must be sought to understand the weight (prevalence) of this norm with respect to the scope of this obligation, balanced with the principle of peace, in contexts of transitions from NIAC to peace. To this end, it is important to respond to some counter-arguments that have been made in the literature and that give little or no weight to Article 6(5) in order to fulfil the obligation to investigate, prosecute and punish.

Of course, all this analysis should not lose sight of the fact that IHL has advocated limits to amnesties, which, as this article will show, are generally shared by other international law regimes, such as IHRL and international criminal law. These limits arise from legitimate concerns related to the fight against impunity and the appropriate balance that must exist between achieving peace and guaranteeing victims’ rights, especially the right to access to justice.³⁸ Without losing sight of the fact that there are indeed limits, the following is a critical response to positions that seek to extend these limits far beyond the requirements derived from IHL.

Article 6(5) from the most restricted positions: A critical response

Some sectors of the doctrine have proposed restricted approaches to analyzing Article 6(5) of AP II. This section explains the scope of these approaches and provides a critical response to them. There are two restricted positions on the interpretation of Article 6(5): the first argues that this article does not contain any international obligation, and the second argues that the provision is limited to “combatant immunity”.

First position: Article 6(5) does not contain any international obligation

The ICRC has clarified that States are not obliged to grant an amnesty at the end of hostilities, but that they do have an obligation to carefully consider this measure and to endeavour to adopt it.³⁹ In this sense, as anticipated above, it could be reasonably interpreted that although there is no international obligation to adopt an amnesty, Article 6(5) does imply an obligation to act, by ensuring that the amnesty is approved.

However, some authors consider that this obligation does not really exist and that amnesty is a matter for States to consider, on a completely optional basis. For example, Mégret acknowledges that the scope of amnesties remains an ambiguous issue in international law. He notes in his critique that

it is very hard to see why states would want to recognize a posteriori the sort of privilege of belligerency that they were not willing to recognize during conflict, except that states were somehow “tricked” by Protocol II into committing themselves to granting amnesties that they would not normally consider in their interest (the record does not attest to this and, therefore one is tempted to treat Article 6.5 as not very serious or representative of states’ real views on the question).⁴⁰

This argument is also related to the argument that Article 6(5) goes beyond the scope of IHL, because amnesties do not pursue any “humanitarian” purpose.⁴¹ In this sense, the discussion on the optional or mandatory scope of Article 6(5) is directly related to the debate on the purposes pursued by the provision. Indeed, it would be necessary to determine whether the rule was intended to protect States’ own interests in the context of the end of hostilities, or whether there is a sufficient basis for considering that it is intended to protect broader purposes and that, therefore, the obligations arising from the rule could be rendered meaningless by reviewing exclusively the “real views” of States.

To respond to these positions, first, given that more generally international law for the protection of the human person is aimed at establishing norms limiting the powers of States, which must yield in order to protect the rights of the human person,⁴² it would be, at the very least, problematic to interpret the scope of the norms in light of States’ self-interest or “real views”. If such an interpretation were made, it could be interpreted that the AP II regulations do not generally contain mandatory legal obligations, which is, of course, an unacceptable premise. In fact, the preamble to the Protocol – which can be used to interpret the treaty provisions⁴³ – states that it protects the humanitarian principles endorsed by common Article 3, which “constitute the foundation of respect for the human person”.⁴⁴ In addition, Article 6 is aimed in its entirety at limiting the punitive power of the State through the norms of due process.

Second, in relation to the regulation's broader purposes, it is clear from various United Nations (UN) pronouncements and specialized doctrine that the purposes of Article 6(5) include (1) ending conflict and fostering reconciliation so that society can move forward,⁴⁵ and (2) facilitating the reintegration of groups that opposed the government during the conflict.⁴⁶ At least the first aim is also supported by the ICRC Commentary to AP II, which notes that the objective of Article 6(5) is “to encourage gestures of reconciliation which can contribute to re-establishing normal relations in the life of a nation which has been divided”.⁴⁷

Likewise, the 2004 report of the UN Secretary-General on the rule of law and transitional justice in conflict and post-conflict societies supports this position, stating that “[c]arefully crafted amnesties can help in the return and reintegration of [displaced civilians and former fighters] and should be encouraged”.⁴⁸ In the same vein, the report of the Secretary-General on the establishment of a Special Court for Sierra Leone stated that “amnesty is an accepted legal concept and a gesture of peace and reconciliation at the end of a civil war or an internal armed conflict”.⁴⁹ This position has also been confirmed in UN practice, which has supported the establishment of amnesties in several conflicts around the world.⁵⁰ The European Union, for its part, has stated that amnesties at the end of hostilities contribute to reconciliation.⁵¹ Finally, the African Union has considered the end of hostilities, the prevention of violence, and reconciliation as purposes of amnesties.⁵²

In sum, many sources agree that there is a broader purpose behind Article 6(5) of AP II: contributing to reconciliation. This end does not depend on States’ real interests or views, but points to principles that transcend those interests. Ultimately, achieving reconciliation is closely related to the broader goal of achieving peace, as a cross-cutting goal of IHL. Therefore, any interpretation of Article 6(5) must be driven by the broader purposes pursued by the provision, rather than by the interests of those who signed and ratified the treaty.

Second position: Article 6(5) is limited to “combatant immunity”

The scope of the conduct included in Article 6(5) has been widely debated. Some point out that while Article 6(5) may include deaths in combat, it does not include the actual act of taking up arms, since there is no jus ad bellum in NIACs.⁵³ In the same vein, other authors affirm that the amnesties encouraged by Article 6(5) refer exclusively to political crimes or crimes directly related to hostilities.⁵⁴ For example, in 1995, Toni Pfanner, head of the ICRC's Legal Division, referred to the relationship between combatant immunity and amnesty, stating:

Article 6(5) of Protocol II is the only and very limited equivalent in the law of non-international armed conflict of what is known in the law of armed conflict as “combatant immunity”, i.e., the fact that a combatant may not be punished for acts of hostility, including killing enemy combatants, as long as he respected international humanitarian law …. [This is given that] [i]n non-international armed conflicts, … those who fight may be punished, under national legislation, for the mere fact of having fought, even if they respected international humanitarian law.⁵⁵

Judge Luz del Carmen Ibáñez Carranza of the ICC included this assessment as support for her argument in her concurring vote of the Appeals Chamber in the Gaddafi case, arguing that with respect to amnesties, the prevailing criterion of IHRL should always apply, although she warned that there may be more flexibility in scenarios of political transition.⁵⁶

However, the ICC Appeals Chamber took a more cautious approach, considering that since Libya's position was clear on the inapplicability of the Amnesty Law in the Gaddafi case, including for failure to demonstrate that the conditions for applying the amnesty had been met,⁵⁷ it was not necessary to analyze the compatibility or incompatibility of the amnesty provisions with international law. It held that the Pre-Trial Chamber's considerations on this issue were obiter dicta and that international law “is still in the developmental stage on the question of acceptability of amnesties”.⁵⁸

In this regard, it should be noted that not only is there not very broad support for interpretations that attempt to limit the conduct included in Article 6(5) to combatant immunity, but there is also significant normative evidence that the article includes a variety of conduct that is not limited to that which pertains to hostilities. The first piece of evidence in this regard derives from Article 6 of AP II itself, which establishes in paragraph 1 that it applies to “the prosecution and punishment of criminal offences related to the armed conflict”. This scope of application, which also extends to paragraph 5 of the same article, is very different and much broader in its wording than regulations that refer to combatant immunity. In fact, both Hague and Geneva law, when referring to combatant immunity, use the expression “participation in hostilities”.⁵⁹ This is why the ICRC has explained that this status means that combatants and members of armed groups cannot “be punished for having committed acts of hostility”.⁶⁰

Thus, if Article 6(1) or 6(5) were intended to restrict application of the article to combatant immunity, the expression “granting amnesty for having committed acts of hostility or for conduct directly related to participation in hostilities” would have been used. However, the regulation uses a broader expression that may include any criminal offence in relation to armed conflict, which would include (under a literal interpretation of the regulation) possible violations of IHL.

The second piece of evidence is that such a narrow interpretation is inconsistent with a contextual interpretation and the spirit of the regulation. While IHL has insisted that permissible acts of hostility per se should not be criminalized,⁶¹ it would be inconsistent for Article 6(5) to empower or invite the adoption of amnesties, rather than simply indicating that the crime itself should not exist. In addition, the provision may be intended to prohibit conduct and not to regulate criminal guarantees.

The third piece of evidence is that the text does not establish which crimes can be included in amnesties; on the contrary, the ICRC interpretation has opted to include only exclusion clauses, leaving national authorities with significant discretion to define the applicable crimes.⁶² In fact, this interpretation of the focus on exclusionary rather than inclusionary criteria is also derived from several UN pronouncements that emphasize the limits to amnesties but not the conduct that could or should be included.⁶³

The fourth piece of evidence is that interpretations which attempt to limit the conduct included in Article 6(5) to combatant immunity completely omit the fact that the ICRC itself has said that combatant status is specific to international armed conflicts and not to NIACs.⁶⁴ In this sense, it is illogical to apply this status in relation to a regulation that was issued to exclusively regulate NIACs.

The last piece of evidence is that, as the UN International Law Commission reiterated in 2018,⁶⁵ there is no international treaty prohibiting the granting of amnesties for any type of conduct, not even international crimes.⁶⁶ However, in Rules 158 and 159 of the ICRC Customary Law Study, the ICRC concluded that customary law prohibits the granting of amnesties for war crimes.⁶⁷ If Article 6 were limited to combatant immunity, the limit would have had to be extended, or it could even have been made clear in the customary rules that the provision was limited to this restricted type of immunity.

Therefore, there are no compelling reasons to maintain that Article 6(5) is limited to combatant immunity. Indeed, as stated by the ICC Appeals Chamber in the Gaddafi case,⁶⁸ the broader question of what is or is not permissible in terms of amnesties has not been fully defined in international law.

There is still significant room for international law to expand the scope of amnesties, and I agree that States’ continued use of amnesties (during the last three decades, at least eighteen countries have used them as part of 200 peace agreements⁶⁹) and the reluctance to codify the prohibition of their use minimize any serious assertion that a rule against amnesty has become a consolidated rule in international law.⁷⁰ However, I do not subscribe to the positions that claim there are no limits on granting amnesties under international law, not even as an emerging custom. According to some of these positions, there is no obligation under international law to investigate, prosecute and punish offenders.⁷¹ On the contrary, there is ample evidence that, in IHRL, IHL and international criminal law, there is a minimum and inviolable prohibition that amnesties may not be granted, under any circumstances, in respect of those most responsible for international crimes.⁷²

Now, while it is true that this inviolable limit exists, some trends in international law have excessively restricted the use of amnesties, without sufficient consideration of the context at the end of hostilities or the effect that Article 6(5) may have on the interpretation of IHRL. This is the case with regard to the IACtHR, which will be analyzed in the following section. Its jurisprudence has a significant impact on Latin America, a region that continues to experience numerous situations of violence, some of which have been classified as NIACs.⁷³

Application of Article 6(5) by the Inter-American Court of Human Rights: Ambiguities and maximalist positions

Within the Inter-American System, the American Convention on Human Rights does not expressly enshrine peace as one of its purposes,⁷⁴ but instead includes specific provisions related to access to justice,⁷⁵ a matter that has been strengthened by the IACtHR's broad jurisprudence regarding the fight against impunity.⁷⁶ In particular, since its first ruling in 1987, in the case of Velásquez Rodríguez v. Honduras, the IACtHR has interpreted the American Convention as containing a specific duty to investigate, prosecute and punish, even though this is not expressly stated in the text of the treaty.⁷⁷ This interpretation was reinforced after the fall of the dictatorial regimes in the Southern Cone and the process of re-establishing democracy in Peru,⁷⁸ Chile,⁷⁹ Uruguay⁸⁰ and Brazil.⁸¹ On the other hand, apart from very tangentially in the case of El Salvador, the IACtHR's jurisprudence has not considered peace as a value, aim or relevant objective in analyzing the scope of the rights contained in the inter-American human rights treaties.

This does not mean that the weight which the IACtHR has given to justice is not justified because, as an important sector of the literature points out, due to the context of institutionalized impunity, for a long time States did not prioritize victims’ rights or international obligations in the framework of the transitions taking place in Latin America.⁸² Therefore, in Latin American contexts, justice has been one of the main demands made by victims in the region⁸³ due to the countries’ lack of a satisfactory response to the call to tackle impunity.⁸⁴ Other regions in the world also share this concern.⁸⁵

Although there are different positions on how the IACtHR has handled amnesties in its jurisprudence,⁸⁶ and there is even a common belief that the jurisprudence is very clear regarding the limits for granting amnesties, self-amnesties and analogous instruments, a careful reading of the jurisprudence reveals – as will be shown – a line of jurisprudence that is inconsistent and ambiguous. This line pays little attention to the context, considering that the prohibitions on this issue come from peremptory norms derived from States’ human rights obligations, which should not give rise to flexibilities. As will be seen, although it would appear that the jurisprudence had a turnaround in recognizing the importance of Article 6(5) of AP II in transitions from armed conflict to peace, this shift ended up being inconsistent and confusing, shedding little light on the future of jurisprudence regarding this issue.

In effect, with respect to conduct for which amnesties are not appropriate, the IACtHR's line of jurisprudence – which favoured limiting amnestiable conduct to that which does not constitute serious human rights violations⁸⁷ – took a supposed turn with the first case that the Court heard regarding a transition from armed conflict to a negotiated peace. Thus, in the Massacres of El Mozote v. El Salvador judgment, the Court cited Article 6(5), together with Rule 159 of the ICRC Customary Law Study in a more permissive sense, insofar as it stated that amnesties could be permitted in contexts of transition from armed conflict to peace, except for those responsible for committing war crimes and crimes against humanity.⁸⁸ In particular, Article 6(5) was not cited in the El Mozote case as a further development in IHRL on the limit of amnesties, as happened in the Gelman case, for example.⁸⁹ Rather, it was cited as the basis⁹⁰ for a potential change, or at least a contextualization, of the previous jurisprudence of the IACtHR.

This contrasts with the Rochac Hernández v. El Salvador judgment, just two years later. Addressing the same Amnesty Law in the same context, the IACtHR returned to the rule that had developed since the Barrios Altos v. Peru case for transitions from dictatorship to democracy, considering the norms of IHL completely irrelevant to analyzing the limits of amnesties.⁹¹ The Court later adopted this same criterion in the case of Guatemala, although not through a judgment but through resolutions on compliance with the judgment issued with respect to this country, which also underwent a transition from armed conflict to a negotiated peace.⁹² There are no explicit reasons to explain this new turn taken by the Court, although the departure of some judges who led the discussion in the El Mozote case or the criticism of Colombia's legal framework for peace by regional organizations recognized in the Inter-American System, such as Human Rights Watch⁹³ and the Colombian Commission of Jurists,⁹⁴ may partly explain it.

More recently, in 2018, in the case of Herzog v. Brazil, the IACtHR reiterated the importance of a NIAC as a criterion for analyzing the scope of amnesties. Although the Court found that the context in Brazil did not warrant the application of IHL (as Brazil was not undergoing a NIAC), it revisited its reasoning from the El Mozote case. This reasoning was entirely overlooked in the Rochac case, which did involve the context of a NIAC. Indeed, in Herzog, the Court stated:

In this regard, it is important to point out that, as this Court has established, international humanitarian law justifies the granting of amnesty laws at the cessation of hostilities in armed conflicts of a non-international nature to enable the return to peace, as long as they do not cover war crimes and crimes against humanity, which cannot remain in impunity.⁹⁵

In fact, the Court later reiterated that violation of Articles 8 and 25 of the American Convention on Human Rights was further exacerbated because “since its approval, the Brazilian Amnesty Law refers to crimes committed outside a non-international armed conflict”.⁹⁶

The IACtHR's most recent opportunity to refer to the impact of IHL on models of transitional justice was the case of Members and Militants of the Patriotic Union v. Colombia. However, although a significant portion of the parties’ arguments focused on the space that the State had in terms of selecting and prioritizing certain conduct and certain responsible parties in the transition from armed conflict to peace, the Court limited itself to giving very tentative support to the transitional justice model in Colombia, without delving into the issue of amnesties.⁹⁷

In short, the IACtHR's jurisprudence regarding the limits of amnesties has been ambiguous and inconsistent in some legal extremes that may be very relevant in designing transitional justice mechanisms. There is still no clear line regarding possible differences in the scope of the obligation of IHL in transitions from armed conflict to peace and the weight it may have, and in particular regarding Article 6(5) of AP II in these contexts, whether or not they have been accompanied by political negotiation. As noted above, the uncertainty surrounding these rules is significant, given the Court's enormous influence,⁹⁸ particularly through concepts such as reviewing compliance,⁹⁹ in a region that will continue to experience several transition processes from armed conflict to peace. The Colombian case, which will be put to the test before the Inter-American System in the near future, will shed important light on this discussion, as will be shown in the following section.

Colombia as a case study: The role of IHL in the framework of peace agreements

In November 2016, after more than four years of negotiation,¹⁰⁰ Colombians’ rejection at the ballot box of the Peace Agreement signed two months earlier between the national government and the Revolutionary Armed Forces of Colombia – People's Army (Fuerzas Armadas Revolucionarias de Colombia – Ejército del Pueblo, FARC),¹⁰¹ and a renegotiation process,¹⁰² the oldest guerrilla group in Latin America and the Colombian government signed the Final Agreement for Ending the Conflict and Building a Stable and Lasting Peace (Final Agreement). Its six central points, derived from the initial framework agenda of the negotiations, were as follows: (1) comprehensive rural reform; (2) guarantees for political participation; (3) mechanisms for demobilization and transition to the end of the conflict; (4) fight against drugs; (5) the creation of a comprehensive system for truth, justice, reparation and guarantees of non-recurrence (pillar of justice model); and (6) implementation, verification and follow-up of the contents of the Agreement.

Regarding the legal nature of the Final Agreement, the parties expressly established that “[t]his Final Agreement to End the Armed Conflict and Build a Stable and Lasting Peace is signed by the National Government and the [FARC-EP], as a Special Agreement pursuant to Article 3, common to the 1949 Geneva Conventions, as per its international standing”.¹⁰³ After the renegotiation, the parties reiterated that this qualification would be given only “as per [common Article 3's] international standing” – in other words, a negotiation to end the conflict in accordance with the parameters of IHL.¹⁰⁴ This qualification as a special agreement of IHL was recognized in a constitutional norm that was later repealed.¹⁰⁵ Likewise, the Constitutional Court declared itself unable (due to lack of jurisdiction) to rule on the legal scope that this denomination could have¹⁰⁶ and qualified the Agreement as an act of a “political nature” or one that was not automatically incorporated into the Constitution, as it lacked a normative nature of its own.¹⁰⁷

However, this same decision clarified that, although the Final Agreement does not have a normative value per se, the contents “that correspond to norms of international humanitarian law and fundamental rights enshrined in the Constitution will be parameters for interpreting, and benchmarks for developing and validating, the norms for implementing the Agreement”.¹⁰⁸

This prominence of IHL had important effects on the justice model. As part of the last peace negotiations with the FARC-EP guerrillas, the justice model began to be widely discussed, long before the Final Agreement was signed. Thus, in 2012, a constitutional reform entitled the Legal Framework for Peace was approved, which included specific legal instruments for transitional justice in the Constitution. Among other issues, it was established that the prioritization and selection criteria were “inherent to the instruments of transitional justice”, so that “efforts could be focused on the criminal investigation into those most responsible for all crimes that acquire the connotation of crimes against humanity, genocide, or war crimes committed systematically” and “the conditional waiver of criminal prosecution of all cases not selected could be authorized”.¹⁰⁹

This constitutional norm led to a series of rulings by the Colombian Constitutional Court related to the justice model in the framework of the transition from armed conflict to peace.¹¹⁰ In a first landmark ruling on the subject, the Constitutional Court analyzed whether this justice model was incompatible with the obligation to investigate, prosecute and punish. The Court concluded that said norm was aligned with constitutional principles, and in its analysis included an interpretation of the difference between granting amnesties in transitions from dictatorship to democracy and in transitions from armed conflict to peace. This interpretation was based on Article 6(5) of AP II and the jurisprudential change in the IACtHR's Massacres of El Mozote case, which the national court classified as an “emblematic case”.¹¹¹ Based on the interpretation of these sources, the Constitutional Court expressly considered that the limit to the amnesties of Article 6(5) would be crimes against humanity and war crimes,¹¹² particularly taking into consideration that in contexts of armed conflict, the rules of IHL are a special rule (lex specialis) with respect to IHRL.¹¹³

Also, in the framework of the normative discussions for implementing the Final Agreement, the law of “amnesty, pardon and special penal treatments” was approved,¹¹⁴ which the Constitutional Court declared as constitutional. In this ruling, the Constitutional Court stated that the obligation to investigate is non-negotiable in the case of war crimes, crimes against humanity and genocide under the terms of the Rome Statute.¹¹⁵

Colombia's transitional justice model is undoubtedly the most complex and sophisticated in Latin America. One of the reasons for this is that the Colombian context (like most societies with protracted armed conflicts) is not only unique but also has extremely complex circumstances, such as (1) the maintenance of strong institutions and the rule of law, despite suffering several internal armed conflicts for more than sixty years (the ICRC even considers that there are currently eight internal armed conflicts in Colombia¹¹⁶); (2) the existence of several illegal armed groups operating in the territory; (3) a transition process without a change in the Political Constitution; and (4) the existence of more than 9 million victims, making this a completely unprecedented scenario of massive human rights violations.

In other academic texts, the present author and others have extensively defended this model's compliance with the State's international obligations.¹¹⁷ On the one hand, there is a prohibition on granting amnesties for the commission of international crimes. On the other hand, although there is the possibility of selecting those most responsible, this is accompanied by a robust “conditionality regime” – a model of investigation of macro-criminal patterns that uncovers complex criminal structures and a robust system of victim participation in the judicial and extrajudicial mechanisms that make up the system, under a restorative justice paradigm.

In effect, in the Colombian model, the conditionality regime meets the conditions for granting penal advantages. This regime has been recognized as a central principle,¹¹⁸ since it reconciles the State's power – in accordance with IHL – to grant penal advantages in the peacebuilding process with its duty to protect victims’ rights to truth, justice, reparation and non-recurrence, mainly derived from IHRL.¹¹⁹

In the Colombian model, those who receive special penal treatment must¹²⁰ provide full truth, guarantee non-recurrence, refrain from committing new crimes, contribute to reparations for victims and provide information on missing persons. In the case of former FARC-EP members, granting penal benefits will also be subject to compliance with the conditions¹²¹ of laying down arms, actively helping to ensure the success of the process of comprehensive reintegration into civilian life, and surrendering minors.

Additionally, restorative justice is one of the guiding paradigms of the judicial component of the transitional justice model in Colombia, the Special Jurisdiction for Peace.¹²² This means that the conditions established for the beneficiaries of mechanisms to streamline criminal prosecution,¹²³ and the sanctions imposed on those convicted,¹²⁴ should adopt a restorative logic, as opposed to an exclusively retributive logic. This principle aims at re-establishing the social pact,¹²⁵ repairing broken social relations,¹²⁶ restoring trust among citizens¹²⁷ and between citizens and institutions, and strengthening a reconciliation scenario.¹²⁸ A dialogic process is therefore encouraged, in which victims actively participate,¹²⁹ as is the creation of a sanctioning regime with a restorative perspective (with non-custodial sanctions),¹³⁰ among other mechanisms.

In support of this model, James Stewart, Deputy Prosecutor of the ICC, stated in 2018 that the mechanisms for streamlining criminal action conceived by Colombia were “consistent with the duty of States to investigate and prosecute the most serious crimes, including war crimes, as an established principle of international law”.¹³¹ As noted above, it is impossible to analyze the obligation to investigate, prosecute and punish without taking into account that judicial truth has considerable limits. Indeed, the defendant has very little incentive to contribute to the truth in an ordinary trial, and therefore clarification of the truth cannot depend exclusively on criminal justice.¹³² It is thus important to create incentives within the judicial process itself and to establish complementary extrajudicial mechanisms.

In fact, the Colombian model has made important advances in matters that had been under investigation within the country's ordinary justice system for several decades,¹³³ which the international community has celebrated and highlighted.¹³⁴ For example, the UN Security Council has described former FARC-EP commanders’ acknowledgement of responsibility as a milestone, after they apologized and pledged to help locate the remains of kidnapping victims who were killed or who died in captivity.¹³⁵ Likewise, the ICC Prosecutor highlighted the results of the Special Jurisdiction for Peace, especially the acknowledgements of responsibility by those most responsible for the conduct, indicating that “[w]hoever wants to turn the page, [by] accepting their responsibility will allow the country to heal and move forward and ensure justice for the victims”.¹³⁶ The Prosecutor also highlighted the recent acknowledgement of responsibility by twenty-one members of the security forces and by participants of the 34th Front of the FARC-EP for the massacre in Bojayá.¹³⁷

Thus, the Colombian model shows that it is possible to implement Article 6(5) of AP II without sacrificing victims’ rights, and indeed to do so while enhancing them. This model therefore represents an opportunity to redefine what is understood in international law as the “fight against impunity”. With respect to this proposed new vision, the Colombian Constitutional Court noted that the Special Jurisdiction for Peace's criminal policy model is the best way to guarantee victims’ rights to truth, justice and reparation. This is because the case selection tool (1) enables efficient use of State resources, (2) is aimed at guaranteeing access to justice for victims within a reasonable time period, (3) allows the investigation to be directed towards establishing a general understanding of the macro context, (4) promotes an exercise of accountability under objective, public and transparent criteria,¹³⁸ and (5) helps overcome the tensions between justice and peace. At the same time, the Court emphasized that access to, and maintenance of, all benefits will be subject to satisfying victims’ rights to truth, reparation and non-recurrence.¹³⁹

The closing of the preliminary examination of Colombia by the ICC Prosecutor on 28 October 2021 provided an important confirmation of the ICC Office of the Prosecutor's support for the new transitional justice model. In its considerations, the Office of the Prosecutor highlighted the Special Jurisdiction for Peace's recent achievements in relation to “pursuing the objectives of retribution, rehabilitation, restoration and deterrence” and took into consideration that “domestic accountability processes are not yet completed, nor all the sentences enforced”.¹⁴⁰ Under the Cooperation Agreement between the Office of the Prosecutor of the ICC and the Government of Colombia, the Office of the Prosecutor committed to “continue supporting Colombia's accountability efforts within its mandate and means”.¹⁴¹

In conclusion

There is no doubt that justice and peace are cross-cutting principles of IHL, and that some of the regulations of this international law regime transcend hostilities to focus on what happens after they have ended. In particular, Article 6(5) of AP II plays a fundamental role in this jus post bellum, since it encourages the use of amnesties to the broadest possible extent, and this should have a necessary impact on States’ obligation to investigate, prosecute and punish, which is derived from IHRL.

The suggestion made by this article is that, in transitions from armed conflict to peace, discussions around Article 6(5) of AP II should stop equating amnesties with impunity schemes. In this sense, the proposal is that Article 6(5) should have a determining effect on the flexibility of the analysis of States’ duty to investigate, prosecute and punish, so that it does not become an “all-or-nothing” rule. This is especially true when ordinary criminal justice has proven ineffective in achieving the objectives pursued and guaranteeing victims’ rights in contexts of massive human rights violations and breaches of IHL. Amnesties, accompanied by a robust conditionality regime, are an instrument that allows us to rethink ways of fighting impunity, as they give more weight to investigating patterns, clarifying complex criminal structures and establishing more holistic and comprehensive systems complemented by extrajudicial mechanisms, which can produce better results and better guarantee victims’ rights.

Undoubtedly, the case of Colombia will represent an important parameter for possible changes in the jurisprudence of various international bodies, including the IACtHR in particular, which until now has been characterized by inflexible approaches to the duty to investigate, prosecute and punish. This change in jurisprudence is urgently needed, considering that the region will still have to experience several transitions from armed conflict to peace, and it is imperative that the demands of justice be balanced with the reconciliation of societies.